Mounting evidence on the prevalence and persistence of early childhood behavior problems, particularly among low-income children of color has fueled interest in early identification and treatment of children at risk for antisocial behavior. However, the instruments typically used to measure behavior problems may have questionable validity among low-income children of color. This study will evaluate the validity of the two most commonly used measures of behavior problems in children 2-4 years old: the Child Behavior Checklist (CBCL) and Eyberg Child Behavior Inventory (ECBF). The aim of Phase 1 is to conduct item analyses on the CBCL and ECBI to examine differential item functioning (DIF) by parent race/ethnicity and household income. African-American, Latino, and White parents (n=840) stratified by income (low vs. middle) and language (monolingual vs. bilingual) will be recruited to complete research measures using a purposive and referral sampling design. Exploratory Item Analysis will be used to test whether items function differently in different racial/ethnic and income groups. Differences on item functioning will be further explored to examine the impact of mediator variables that might explain racial/ethnic or income differences. The aims of Phase 2 are to (a) describe and compare the acceptability of the ECBI and CBCL among African-American, Latino, and White parents from low and middle-income groups and (b) understand how significant item differences obtained in Phase I might be related to parents' race/ethnicity, income group, and/or language. Data will be obtained from 21 focus groups of parents of 2-4 year olds stratified by race/ethnicity, income, and language. Themes will be extracted and analyzed using qualitative methods. The aim of Phase 3 is to integrate the findings of Phases I and 2 and generate recommendations for using and revising these measures so they may provide more valid estimates of behavior problems among young children of color.